🔬 MyHOME v2.0.0b17 Pre-Release: in traces we find the truth #856
Replies: 1 comment
|
First of all, thank you for all the work on MyHOME, and congratulations on the v2.0.0b17 pre-release. A status note from the OWNd / MyHOME audit side, for the record here. The fixes now in flight are the last protocol-level findings. The file-by-file audit of OWNd and MyHOME was cross-checked three ways, DeepSeek, Kimi and Claude, against the official WHO documents and the firmware oracle, and the few points where the reviewers disagreed were settled on the F461 test plant rather than argued. What survived that check is being fixed in the open PRs: a WHO 7 camera WHERE zero-pad, a send-layer guard for the WHO 18 dimension 52 four-digit-year frame that crashes F460/F461, WHO 14 General and Area addressing per the WHO 1 PDF table, a WHO 13 note on the dimensions 10/11/12 that crash H4684/MH200, and the bare WHO 22 power frame that crashes the sound daemon on F461/F454 (the with-source form goes out clean). After this round we do not expect further findings that come from misreading OpenWebNet. Whatever is left is integration logic, cover calibration being the obvious example, not a misinterpretation of the protocol. Thanks again for making this a repository where the traces decide. |
Uh oh!
There was an error while loading. Please reload this page.
🔬 MyHOME v2.0.0b17 is out: in traces we find the truth
OpenWebNet-HA/MyHOME v2.0.0b17 is tagged and ready for field testing: 61 pull requests since b16, on top of the new OWNd 2.0.0b13.
📦 Release notes and
myhome.zip: v2.0.0b17📚 Docs: openwebnet-ha.github.io/MyHOME/beta/
📡 The theme of b17: "In traces we find the truth"
When working with home automation protocols—and OpenWebNet in particular—official documentation and specifications can only take you so far. Between undocumented firmware nuances across gateway families (MH200N, MH201, F454, MyHomeServer1, Legrand 067292, F461), actuator configurator settings (
M=0,M=1,M=4on F411/F461), and physical motor kinetics, the only ground truth is authentic bus traces captured on live installations.In
v2.0.0b17, our community of testers and trace contributors proved this once again. By providing authentic on-wire captures from physical plants, we were able to pinpoint and eliminate elusive edge cases that could never have been deduced in theory:travel_time + 2 sfrom known positions, and an actuator stop within 0.5 s before stored travel time is recognized as the true limit stop rather than freezing position at 98% or 1%.Every single trace submitted is converted into a deterministic test fixture in CI, ensuring that no future refactor can ever break real-world installations.
🏆 Deep thanks to our reviewers
A release of this magnitude is only possible with reviewers who scrutinize every frame, timing edge case, and architectural boundary:
❤️ And heartfelt thanks to our testers and trace providers
Our deepest gratitude goes to the community members who took the time to capture logs, export traces, and test beta candidates on their living systems:
min_position_delta(🔬 MyHOME v2.0.0b16 Pre-Release: having a firm conversation with the firmware #770, feat(cover): per-cover min_position_delta for advanced covers #778).You make this project what it is. Thank you! 🙏
🌟 What's new in v2.0.0b17
🪟 Covers (WHO 2) & Calibration
travel_time + 2 sfrom known positions: Every full run (open_cover,close_cover, scope runs, wall-switch runs) now settles at 0% / 100% after the stored travel time plus 2 s has elapsed, even when gateways do not relay actuator stops (fix(cover): settle full open/close, scope, and wall runs on travel_time + 2 s (#770) #834).min_position_delta(1–5%, default 2%): Allows fine-grained 1% adjustments for tall blinds and precise slat tilts (feat(cover): per-cover min_position_delta for advanced covers #778).*2*1000#0*WHERE##) halt timed runs: Gateways that relay scenario or keypad stops as translated frames now properly stop timed runs (fix(cover): settle full open/close, scope, and wall runs on travel_time + 2 s (#770) #834).myhome.calibrate_coverrefuses runs ending near whole-minute configurator cutoffs (59–65 s, 119–130 s onF411atM=1, up to 10 min) ([Bug]:calibrate_coverstores the F411 120 s cutoff (M=1) as a travel time #837, fix(cover): refuse every configurator run-time limit in calibrate_cover, not only 60 s (#837) #843).🌡️ Climate & Thermoregulation (WHO 4)
switchentities to manually actuate heating solenoids on demand (Issue climate zone ON/OFF #785, PR feat(switch,climate): support WHO=4 probe-less thermo actuator switches (#785) #817).💡 Lighting & Dimmers (WHO 1)
WHAT 101–199) and autodiscovered dimmers ([Bug]: a coarse WHAT 2..10 status reply overwrites the exact DIMENSION 1 level of a fine dimmer (F418 on F461) #851, fix(light): ask Dimension 1 after a coarse preset echo on a fine dimmer (#851) #852), and aligned coarse preset curve values with hardware dimming levels (fix(light,expose): map coarse WHAT 2..10 presets via BTicino dimmerLevelTo100 curve (#466) #828).WHERE([2.0.0b16] Configured light removed on reload after motion sensor discovery at the same WHO 1 address (F454 / Legrand 067226) #777, fix(light): keep a configured light when a WHO 1 motion sensor shares its address (#777) #781).🔊 Sound System & Burglar Alarm
🔌 Gateways, Engine & Protocol
🛡️ Quality & Test Engine
compat_name, logger compat, deprecation guards).📥 How to install / upgrade
HACS:
Update to v2.0.0b17 (ensure Show beta versions is enabled in integration options).
Manual CLI / SSH:
Requires Home Assistant 2026.3 or newer.
🧪 What we'd love feedback on
Please share your thoughts, test results, and any new traces below. Thank you all for your incredible support and collaboration! 🚀📡
All reactions